00_EDA

Machine Learning

Gonzalo Rodríguez Cañada

Data Sourcing

Content of the diamonds data set

This dataset contains the prices and other attributes of almost 54000 diamonds. Attributes, are explained in the following lines.

Price

Price in US dollars

Carat

Weight of the diamond. 200 mg is equivalent to 1 carat.Price increases with the weight.If 2 diamonds weight are equal, then the other features take part to determine the price.

Cut

Quality of the cut which is based on three parameters. Brilliance,fire and scintillation. Other features of a cut are depht percentage and table percentage.

Depth

Is the distance from the surface to the culet. Total depth percentage = z / mean(x, y) = 2 * z / (x + y)

Table

The width of the diamond's table(the topmost surface) expressed as a percentage of its average diameter.

Color

Color measurement in diamond measures lacks color.This parameter show the presence of color in a diamond. D(best) Colorless -J (worst) Light color

Clarity

A measurement of how clear the diamond is based on the inclusions.Is categorized as (I1 (worst), SI2, SI1, VS2, VS1, VVS2, VVS1, IF (best))

X ,Y,Z

Length, Width and Depth, is the diamond dimension measured in milimeters.

In [1]:
import pandas as pd 
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
from pandas_profiling import ProfileReport
import seaborn as sns
In [31]:
#we create the dataframes with the function pd.read_csv since we start from files in this format

diamonds = pd.read_csv('../data/diamonds.csv',delimiter = ',')  #We indicate that the separator is the comma
In [32]:
diamonds.head(10)
Out[32]:
Unnamed: 0 carat cut color clarity depth table price x y z
0 1 0.23 Ideal E SI2 61.5 55.0 326 3.95 3.98 2.43
1 2 0.21 Premium E SI1 59.8 61.0 326 3.89 3.84 2.31
2 3 0.23 Good E VS1 56.9 65.0 327 4.05 4.07 2.31
3 4 0.29 Premium I VS2 62.4 58.0 334 4.20 4.23 2.63
4 5 0.31 Good J SI2 63.3 58.0 335 4.34 4.35 2.75
5 6 0.24 Very Good J VVS2 62.8 57.0 336 3.94 3.96 2.48
6 7 0.24 Very Good I VVS1 62.3 57.0 336 3.95 3.98 2.47
7 8 0.26 Very Good H SI1 61.9 55.0 337 4.07 4.11 2.53
8 9 0.22 Fair E VS2 65.1 61.0 337 3.87 3.78 2.49
9 10 0.23 Very Good H VS1 59.4 61.0 338 4.00 4.05 2.39
In [33]:
diamonds.tail(10)
Out[33]:
Unnamed: 0 carat cut color clarity depth table price x y z
53930 53931 0.71 Premium E SI1 60.5 55.0 2756 5.79 5.74 3.49
53931 53932 0.71 Premium F SI1 59.8 62.0 2756 5.74 5.73 3.43
53932 53933 0.70 Very Good E VS2 60.5 59.0 2757 5.71 5.76 3.47
53933 53934 0.70 Very Good E VS2 61.2 59.0 2757 5.69 5.72 3.49
53934 53935 0.72 Premium D SI1 62.7 59.0 2757 5.69 5.73 3.58
53935 53936 0.72 Ideal D SI1 60.8 57.0 2757 5.75 5.76 3.50
53936 53937 0.72 Good D SI1 63.1 55.0 2757 5.69 5.75 3.61
53937 53938 0.70 Very Good D SI1 62.8 60.0 2757 5.66 5.68 3.56
53938 53939 0.86 Premium H SI2 61.0 58.0 2757 6.15 6.12 3.74
53939 53940 0.75 Ideal D SI2 62.2 55.0 2757 5.83 5.87 3.64
In [34]:
diamonds.shape #in order to observe the number of rows and columns of our dataset
Out[34]:
(53940, 11)
In [35]:
diamonds.cut.unique() #To know the different type of qualities
Out[35]:
array(['Ideal', 'Premium', 'Good', 'Very Good', 'Fair'], dtype=object)
In [36]:
#Let's get an idea of how many we have of each
diamonds.cut.value_counts()
Out[36]:
Ideal        21551
Premium      13791
Very Good    12082
Good          4906
Fair          1610
Name: cut, dtype: int64

Data Cleaning

In [37]:
diamonds. isnull(). sum() #theres no NaN
Out[37]:
Unnamed: 0    0
carat         0
cut           0
color         0
clarity       0
depth         0
table         0
price         0
x             0
y             0
z             0
dtype: int64
In [38]:
diamonds.iloc[:,0] = diamonds.iloc[:,0].astype('str')#we dont want the index to be in the summary so we conver in object
 # a summary of the numeric columns to have an idea of the range of the values.
In [39]:
duplicated_values = diamonds[diamonds.duplicated()]
print(duplicated_values) #theres no duplicated values

Empty DataFrame
Columns: [Unnamed: 0, carat, cut, color, clarity, depth, table, price, x, y, z]
Index: []
In [40]:
diamonds1 = diamonds.rename(columns={'Unnamed: 0': 'Index','z':'depth_mm','x':'length_mm','y':'width_mm','depth':'depth_%'})
# Changing the name of the column.
In [41]:
diamonds1.info() #in order to know the Dtype, and confirming Index has an object Dtype

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 53940 entries, 0 to 53939
Data columns (total 11 columns):
 #   Column     Non-Null Count  Dtype  
---  ------     --------------  -----  
 0   Index      53940 non-null  object 
 1   carat      53940 non-null  float64
 2   cut        53940 non-null  object 
 3   color      53940 non-null  object 
 4   clarity    53940 non-null  object 
 5   depth_%    53940 non-null  float64
 6   table      53940 non-null  float64
 7   price      53940 non-null  int64  
 8   length_mm  53940 non-null  float64
 9   width_mm   53940 non-null  float64
 10  depth_mm   53940 non-null  float64
dtypes: float64(6), int64(1), object(4)
memory usage: 4.5+ MB
In [42]:
diamonds1.describe() #Here we can see a quick summary of the numercical features
Out[42]:
carat depth_% table price length_mm width_mm depth_mm
count 53940.000000 53940.000000 53940.000000 53940.000000 53940.000000 53940.000000 53940.000000
mean 0.797940 61.749405 57.457184 3932.799722 5.731157 5.734526 3.538734
std 0.474011 1.432621 2.234491 3989.439738 1.121761 1.142135 0.705699
min 0.200000 43.000000 43.000000 326.000000 0.000000 0.000000 0.000000
25% 0.400000 61.000000 56.000000 950.000000 4.710000 4.720000 2.910000
50% 0.700000 61.800000 57.000000 2401.000000 5.700000 5.710000 3.530000
75% 1.040000 62.500000 59.000000 5324.250000 6.540000 6.540000 4.040000
max 5.010000 79.000000 95.000000 18823.000000 10.740000 58.900000 31.800000

Data Visualization

Numerical features

Correlation matrix

In [43]:
corr = diamonds1.corr()
# Generate a mask for the upper triangle
mask = np.triu(np.ones_like(corr, dtype=bool))

# Set up the matplotlib figure
f,ax = plt.subplots(figsize=(15,15))

# Generate a custom diverging colormap
cmap = sns.diverging_palette(230, 20, as_cmap=True)

# Draw the heatmap with the mask and correct aspect ratio
sns.heatmap(corr, mask=mask, cmap=cmap, vmax=1, center=0,
            square=True, linewidths=.6, cbar_kws={"shrink": .5})
Out[43]:
<AxesSubplot:>

Correlation matrix shows how our features are related. As is observed,there is a strong correlation between the weight of our diamond with the dimensions and the price, as was commented previously.Additionally, as is expected theres a high correlation between the dimensions of our diamond.

In [44]:
sns.pairplot(diamonds1, hue="cut", diag_kind="hist")
Out[44]:
<seaborn.axisgrid.PairGrid at 0x29a3e480520>

Pair-plots are way to see relationships between each variable. It produces a matrix of relationships between each variable.

Histograms

Let's take a look about how data is distributed as function of the cut which represent the quality of the diamond.

In [45]:
sns.displot(diamonds1, x="carat", hue="cut",kde=True)
plt.xlabel('Weight of the diamond in grams')
Out[45]:
Text(0.5, 6.79999999999999, 'Weight of the diamond in grams')
In [46]:
sns.displot(diamonds1, x="depth_%", hue="cut",kde=True)
plt.xlabel('Depth percentage')
Out[46]:
Text(0.5, 6.79999999999999, 'Depth percentage')
In [47]:
sns.displot(diamonds1, x="table", hue="cut",kde=True)
plt.xlabel('Table percentage')
Out[47]:
Text(0.5, 6.79999999999999, 'Table percentage')
In [48]:
sns.displot(diamonds1, x="price", hue="cut",kde=True)
plt.xlabel('$')
Out[48]:
Text(0.5, 6.79999999999999, '$')
In [49]:
sns.displot(diamonds1, x="length_mm", hue="cut",kde=True)
plt.xlabel('Length in mm')
Out[49]:
Text(0.5, 6.79999999999999, 'Length in mm')
In [50]:
sns.displot(diamonds1, x="width_mm", hue= "cut",kde=True)
plt.xlabel('Width in mm')
Out[50]:
Text(0.5, 6.79999999999999, 'Width in mm')
In [51]:
sns.displot(diamonds1, x="depth_mm", hue="cut",kde=True)
plt.xlabel('Depth in mm')
Out[51]:
Text(0.5, 6.79999999999999, 'Depth in mm')

Categorical features

Boxplots

The boxplots depict groups of numerical data through their quantiles.

In [52]:
sns.boxplot(x='cut',y='price',data=diamonds1)
Out[52]:
<AxesSubplot:xlabel='cut', ylabel='price'>

From this boxplot can be observed that except for ideal cut, the lower is the quality, there are more outliers. Moreover, each type of cut has the same minimum and maximum price.

In [53]:
sns.boxplot(x='clarity',y='price',data=diamonds1)
Out[53]:
<AxesSubplot:xlabel='clarity', ylabel='price'>

No trend can be drawn from this graph, however,each type of cut has the same minimum and maximum price.

In [54]:
sns.boxplot(x='color',y='price',data=diamonds1)
Out[54]:
<AxesSubplot:xlabel='color', ylabel='price'>

From this boxplot can be observed that except for G, the better is the quality of the color, there are more outliers. Moreover, each type of cut has the same minimum and maximum price.

Count plots

In [55]:
cut_clarity_table = pd.crosstab(index=diamonds1["cut"], columns=diamonds1["clarity"])

cut_clarity_table.plot(kind="bar", 
                 figsize=(12,12),
                 stacked=True)
Out[55]:
<AxesSubplot:xlabel='cut'>

As it observed that quality of cut if the parameter that drives people to buy the diamond,regardless the quality.

In [56]:
color_clarity_table = pd.crosstab(index=diamonds1["color"], columns=diamonds1["clarity"])

color_clarity_table.plot(kind="bar", 
                 figsize=(12,12),
                 stacked=True)
Out[56]:
<AxesSubplot:xlabel='color'>

People prefer G and E colour over the others, as well as the the most sought clarities are VS2 and SI1.

In [57]:
clarity_cut_frame = pd.crosstab(index=diamonds1["clarity"], columns=diamonds1["cut"])

clarity_cut_frame.plot(kind="bar", 
                 figsize=(12,12),
                 stacked=True)
Out[57]:
<AxesSubplot:xlabel='clarity'>

The most bought diamonds are the ones with the clarity SI1 and VS2,with an ideal cut, followed closely by the premium cut, nonetheless, it is observed that people preffer to choose by color instead of clarity.

Profile Report

In [58]:
profile = ProfileReport(diamonds1,title='EDA of Diamonds Dataset')
profile










    






 
 















    















    






 
 















    















    
















    
Out[58]:





























References:
Seaborn Graphics-https://seaborn.pydata.org/generated/seaborn.pairplot.html


Profile report-https://towardsdatascience.com/4-libraries-that-can-perform-eda-in-one-line-of-python-code-b13938a06ae


Histograms with seaborn- https://towardsdatascience.com/histograms-and-density-plots-in-python-f6bda88f5ac0


Diamonds dataset source-https://www.kaggle.com/shivam2503/diamonds